Posted by
statisticsdoc on
Sep 24, 2006; 2:39am
URL: http://spssx-discussion.165.s1.nabble.com/2nd-order-factor-analysis-tp1071115p1071116.html
Stephen Brand
www.statisticsdoc.com
Dear Dr. Joshi,
What you describe is a potentially useful approach to exploratory analysis
for the purpose of scale construction. However, additional analyses should
be performed in order to confirm the structure and internal consistency of
the scales. First, as you state in your next e-mail to this list, AMOS (or
your favorite structural equation modeling package) can be utilized to
conduct a confirmatory factor analysis of the proposed five-dimensional
model for scoring the items into scales. In addition, carry out reliability
analyses to examine the internal consistency of the scales. With five
scales and 20 items, some scales might have very few items and hence have
relatively Cronbach alpha levels.
As an aside, the term "secondary factor analysis" is sometimes used in two
quite different - even opposite senses. The most widely accepted use refers
to the analysis of oblique factors to determine the higher order dimensions
underlying the factor structure (e.g., consider a lengthy personality
inventory with 17 first-order oblique factors - the 17 oblique factors might
in turn reflect a smaller set of say 2 to 5 higher order secondary factors.
A quite different sense is when factor analysis is used to split apart a
dimension, as you have done, by seeing whether the items that load on the
same factor in the initial analysis can be differentiated into separate
factors when just those items are factor analzed. This is sometimes a quite
useful analysis when a factor (often the first), combines items that are
correlated but seem to tap different sub-areas of a dimension. The latter
is a useful exporatory approach, but the results stand in need of a
confirmatory analysis.
Best,
Stephen Brand
For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]]On Behalf Of
DR VEENA Joshi
Sent: Saturday, September 23, 2006 11:03 AM
To:
[hidden email]
Subject: 2nd order factor analysis
Hi,
I have responses on Healthcare perceptions on 20 questions (likert scale of
5).
Using PCA Varimax rotation I got 5 factors. They are not correlated.
Is it correct to use factor analysis (second order) for the variables
in the first factor to get 2 factors? (I have read in a book that one
can use 2nd order factor analysis only if one has used oblic
rotation).
After getting 2 factors from the first factor, I now have 6 factors
and some of them are correlated with these two new factors.
Is this corect procedure? If not what will be the correct procedure?
After getting 6 factors I used these factor scores as covariates and
used multinomial regression. I am getting expected results. But not
sure if I have followed correct procedure.
Factor scores should be correlated or not ? When I did PAF varimax,
or obline, factor scores correlated.
Which will be the best model?
Veena